© Neuronpedia 2026
    Privacy & TermsBlogGitHubSlackTwitterContact
    Neuronpedia logo - a computer chip with a rounded viewfinder border around it

    Neuronpedia

    Natural Language
    Autoencoders
    NEW    Assistant AxisNEWCircuit TracerUPDATESteerSAE EvalsExportsAPI Community BlogPrivacy & TermsContact
    1. Home
    3. Joseph Bloom · Initial Gemma-2B-IT SAEs
    5. Gemma-2B-IT
    7. Residual Stream
    9. 12-RES-JB
    11. 1826
    Prev
    Next
    INDEX
    Explanations

    instructions or recipes for making food items, specifically dosa and chutney

    oai_token-act-pair · gpt-3.5-turboTriggered by @bot
    New Auto-Interp
    Top Features by Cosine Similarity
    Configuration
    jbloom/Gemma-2b-IT-Residual-Stream-SAEs/gemma_2b_it_blocks.12.hook_resid_post_16384
    Prompts (Dashboard)
    24,576 prompts, 128 tokens each
    Dataset (Dashboard)
    HuggingFaceFW/fineweb
    Features
    16,384
    Data Type
    float32
    Hook Name
    blocks.12.hook_resid_post
    Hook Layer
    12
    Architecture
    standard
    Context Size
    1,024
    Dataset
    Skylion007/openwebtext
    Activation Function
    relu
    Embeds
    IFrame
    Link
    Not in Any Lists

    No Comments

    Neuron Alignment
    Index
    Value
    % of L₁
    297
    +0.12
    0.3%
    906
    +0.09
    0.3%
    1253
    +0.08
    0.2%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    297
    +0.12
    0.05
    766
    +0.09
    0.04
    924
    +0.08
    0.04
    Negative Logits
    <bos>
    -0.81
     alkoh
    -0.74
     silikon
    -0.74
     makro
    -0.72
     plak
    -0.70
     optik
    -0.70
     antik
    -0.65
     elek
    -0.62
     hecta
    -0.62
     kram
    -0.61
    POSITIVE LOGITS
     perciò
    0.70
     disreg
    0.68
     suscep
    0.64
     poichè
    0.63
     purtroppo
    0.60
     impra
    0.60
     maneu
    0.60
     Varies
    0.60
     esattamente
    0.59
     ovunque
    0.57
    Activations Density 0.364%

    No Known Activations